Module opencv::sfm

Structure From Motion

The opencv_sfm module contains algorithms to perform 3d reconstruction from 2d images.

The core of the module is based on a light version of Libmv originally developed by Sameer Agarwal and Keir Mierle.

Whats is libmv?

libmv, also known as the Library for Multiview Reconstruction (or LMV), is the computer vision backend for Blender’s motion tracking abilities. Unlike other vision libraries with general ambitions, libmv is focused on algorithms for match moving, specifically targeting Blender as the primary customer. Dense reconstruction, reconstruction from unorganized photo collections, image recognition, and other tasks are not a focus of libmv.

Development

libmv is officially under the Blender umbrella, and so is developed on developer.blender.org. The source repository can get checked out independently from Blender.

This module has been originally developed as a project for Google Summer of Code 2012-2015.

Note:

• Notice that it is compiled only when Eigen, GLog and GFlags are correctly installed.

Check installation instructions in the following tutorial: [tutorial_sfm_installation]

Conditioning

Fundamental

Input/Output

Numeric

Projection

Robust Estimation

Triangulation

Reconstruction

Note: - Notice that it is compiled only when Ceres Solver is correctly installed.

      Check installation instructions in the following tutorial: [tutorial_sfm_installation]

Simple Pipeline

Note: - Notice that it is compiled only when Ceres Solver is correctly installed.

       Check installation instructions in the following tutorial: [tutorial_sfm_installation]

Modules

• prelude

Structs

• BaseSFM
  base class BaseSFM declares a common API that would be used in a typical scene reconstruction scenario
• SFMLibmvEuclideanReconstruction
  SFMLibmvEuclideanReconstruction class provides an interface with the Libmv Structure From Motion pipeline.
• libmv_CameraIntrinsicsOptions
  Data structure describing the camera model and its parameters.
• libmv_ReconstructionOptions
  Data structure describing the reconstruction options.

Constants

• SFM_DISTORTION_MODEL_DIVISION
• SFM_DISTORTION_MODEL_POLYNOMIAL
• SFM_IO_BUNDLER
• SFM_IO_OPENMVG
• SFM_IO_OPENSFM
• SFM_IO_THEIASFM
• SFM_IO_VISUALSFM
• SFM_REFINE_FOCAL_LENGTH
• SFM_REFINE_PRINCIPAL_POINT
• SFM_REFINE_RADIAL_DISTORTION_K1
• SFM_REFINE_RADIAL_DISTORTION_K2

Traits

• BaseSFMTrait
  Mutable methods for crate::sfm::BaseSFM
• BaseSFMTraitConst
  Constant methods for crate::sfm::BaseSFM
• SFMLibmvEuclideanReconstructionTrait
  Mutable methods for crate::sfm::SFMLibmvEuclideanReconstruction
• SFMLibmvEuclideanReconstructionTraitConst
  Constant methods for crate::sfm::SFMLibmvEuclideanReconstruction

Functions

• apply_transformation_to_points
  Apply Transformation to points.
• compute_orientation
  Computes Absolute or Exterior Orientation (Pose Estimation) between 2 sets of 3D point.
• depth
  Returns the depth of a point transformed by a rigid transform.
• essential_from_fundamental
  Get Essential matrix from Fundamental and Camera matrices.
• essential_from_rt
  Get Essential matrix from Motion (R’s and t’s ).
• euclidean_to_homogeneous
  Converts points from Euclidean to homogeneous space. E.g., ((x,y)->(x,y,1))
• fundamental_from_correspondences7_point_robust
  Estimate robustly the fundamental matrix between two dataset of 2D point (image coords space).
• fundamental_from_correspondences7_point_robust_def
  Estimate robustly the fundamental matrix between two dataset of 2D point (image coords space).
• fundamental_from_correspondences8_point_robust
  Estimate robustly the fundamental matrix between two dataset of 2D point (image coords space).
• fundamental_from_correspondences8_point_robust_def
  Estimate robustly the fundamental matrix between two dataset of 2D point (image coords space).
• fundamental_from_essential
  Get Essential matrix from Fundamental and Camera matrices.
• fundamental_from_projections
  Get Fundamental matrix from Projection matrices.
• homogeneous_to_euclidean
  Converts point coordinates from homogeneous to euclidean pixel coordinates. E.g., ((x,y,z)->(x/z, y/z))
• import_reconstruction
  Import a reconstruction file.
• import_reconstruction_def
  Import a reconstruction file.
• isotropic_preconditioner_from_points
  Point conditioning (isotropic).
• k_rt_from_projection
  Get K, R and t from projection matrix P, decompose using the RQ decomposition.
• mean_and_variance_along_rows
  Computes the mean and variance of a given matrix along its rows.
• motion_from_essential
  Get Motion (R’s and t’s ) from Essential matrix.
• motion_from_essential_choose_solution
  Choose one of the four possible motion solutions from an essential matrix.
• normalize_fundamental
  Normalizes the Fundamental matrix.
• normalize_isotropic_points
  This function normalizes points. (isotropic).
• normalize_points
  This function normalizes points (non isotropic).
• normalized_eight_point_solver
  Estimate the fundamental matrix between two dataset of 2D point (image coords space).
• preconditioner_from_points
  Point conditioning (non isotropic).
• projection_from_k_rt
  Get projection matrix P from K, R and t.
• projections_from_fundamental
  Get projection matrices from Fundamental matrix
• reconstruct
  Reconstruct 3d points from 2d correspondences while performing autocalibration.
• reconstruct_1
  Reconstruct 3d points from 2d correspondences while performing autocalibration.
• reconstruct_1_def
  Reconstruct 3d points from 2d correspondences while performing autocalibration.
• reconstruct_2
  Reconstruct 3d points from 2d images while performing autocalibration.
• reconstruct_2_def
  Reconstruct 3d points from 2d images while performing autocalibration.
• reconstruct_3
  Reconstruct 3d points from 2d images while performing autocalibration.
• reconstruct_3_def
  Reconstruct 3d points from 2d images while performing autocalibration.
• reconstruct_def
  Reconstruct 3d points from 2d correspondences while performing autocalibration.
• relative_camera_motion
  Computes the relative camera motion between two cameras.
• skew
  Returns the 3x3 skew symmetric matrix of a vector.
• triangulate_points
  Reconstructs bunch of points by triangulation.